Scaffolding idea

[Waethorn]

@Boreas618

In regards to your blog23 article last week, I had an idea a few weeks ago and I think it might line up with what you're discussing. The idea was to use Bloom's taxonomy (the 1956 noun version for better control, and self evaluation for output, rather that the 2001 revision), optionally including the affective domain for the purpose of analyzing user response (like a baked-in emotional safety filter for better interaction), and the psychomotor domain adapted to tool use for agents, while integrating Webb's DOK in the cognitive domain. I don't know of any projects using this explicitly, but my agent interactions seem to provide better outputs with it so far. Any thoughts on this?

[Boreas618]

This is a compelling idea and aligns with the purpose of Scaffolding to enable easy expression and prototyping of agents across diverse architectures/philosophies.